Recognizing Human Falls and Routine Activities Using Accelerometers
Sudarshan S. Chawathe · 2019
Detecting falls and other mishaps using data from sensors worn by individuals is an important task with applications in healthcare. A related task is using such sensor data to detect routine activities of daily living. This paper models such detection of falls and routine activities as a classification problem. Using a publicly available dataset of real accelerometer traces generated by participants performing intentional falls and other activities, the efficacy and performance of several classifiers are studied experimentally.